Image Feature Extraction
Transformers
JAX
Safetensors
MLX
PyTorch
aimv2_vision_model
vision
custom_code
Instructions to use apple/aimv2-large-patch14-native with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use apple/aimv2-large-patch14-native with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-feature-extraction", model="apple/aimv2-large-patch14-native", trust_remote_code=True)# Load model directly from transformers import AutoImageProcessor, AutoModel processor = AutoImageProcessor.from_pretrained("apple/aimv2-large-patch14-native", trust_remote_code=True) model = AutoModel.from_pretrained("apple/aimv2-large-patch14-native", trust_remote_code=True) - MLX
How to use apple/aimv2-large-patch14-native with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir aimv2-large-patch14-native apple/aimv2-large-patch14-native
- Notebooks
- Google Colab
- Kaggle
- Local Apps
- LM Studio
fix: crash because pos enc is on CPU device
#1
by karimknaebel - opened
- modeling_aimv2.py +1 -1
modeling_aimv2.py
CHANGED
|
@@ -101,7 +101,7 @@ class AIMv2ViTPreprocessor(nn.Module):
|
|
| 101 |
tokens = self.patchifier(x)
|
| 102 |
pos_embed = get_sincos_pos_embed(
|
| 103 |
H // self.patch_h, W // self.patch_w, embed_dim=self.embed_dim
|
| 104 |
-
)
|
| 105 |
tokens = tokens + pos_embed
|
| 106 |
return tokens
|
| 107 |
|
|
|
|
| 101 |
tokens = self.patchifier(x)
|
| 102 |
pos_embed = get_sincos_pos_embed(
|
| 103 |
H // self.patch_h, W // self.patch_w, embed_dim=self.embed_dim
|
| 104 |
+
).to(tokens.device)
|
| 105 |
tokens = tokens + pos_embed
|
| 106 |
return tokens
|
| 107 |
|